Opus 4.6 is outstanding for code, and for the little I have used it outside of that context, in everything else I have used it with. The productivity with code is at least 3x what I was getting with 5.2, and it can handle entire projects fairly responsibly. It doesn’t patronize the user, and it makes a very strong effort to capture and follow intentions. Unlike 5.2, I’ve never had to throw out a days work that it covertly screwed up taking shortcuts and just guessing.
I like seeing this analysis on new model releases, any chance you can aggregate your opinions in one place (instead of the hackernews comment sections for these model releases)?